2019 Volume 69 Issue 4 Pages 573-584
The green rice leafhopper (GRH, Nephotettix cincticeps Uhler) is a serious insect pest of rice (Oryza sativa L.) in temperate regions of Asia. Wild Oryza species are the main source of resistance to insects. The W1413 accession of African wild rice (O. longistaminata A. Chev. & Roehrich) is resistant to GRH. To analyze its resistance, we developed 28 BC3F3 introgression lines carrying W1413 segments in the genetic background of Nipponbare, a susceptible rice cultivar, and evaluated their GRH resistance. Five BC3F3 populations were used for quantitative trait locus (QTL) analysis and seven BC3F4 populations for QTL validation. Four significant QTLs on the long arm of chromosome 2 (qGRH2), short arm of chromosome 4 (qGRH4), short arm of chromosome 5 (qGRH5), and long arm of chromosome 11 (qGRH11) were identified. The contribution of the W1413 allele at qGRH11 was the largest among the four QTLs; the other QTLs also contributed to GRH resistance. Chromosomal locations suggested that qGRH11 corresponds to the previously reported GRH resistance gene Grh2, qGRH4 to Grh6, and qGRH5 to Grh1. qGRH2 is a novel QTL for resistance to GRH. Thus, resistance of O. longistaminata to GRH can be explained by at least four QTLs.
Rice (Oryza sativa L.) is the most important crop and essential daily food worldwide (Khush 2003). Wild rice is a rare germplasm resource and very important for biodiversity conservation (Zhang and Xie 2014) . The AA-genome species of the genus Oryza have valuable genes for the improvement of cultivated rice, in particular genes conferring resistance to brown planthopper (BPH), small brown planthopper, white-backed planthopper, green leafhopper (GLH), green rice leafhopper (GRH), yellow stem borer, leaf and neck blast, bacterial leaf blight, and sheath blight (Brar and Khush 1997, Fujita et al. 2013, Jena 2010, Khush 1997).
Host-plant resistance has been used to reduce direct damage to plants from insect attack and indirect damage from viral transmission (Brar et al. 2009). Using resistant varieties helps to manage the problems caused by insect pests such as GRH, GLH, and BPH. One of the most serious insect pests of rice, GRH (Nephotettix cincticeps Uhler), is distributed mostly in the temperate regions of East Asia (Ghauri 1971) and is widely distributed in Japan (Hokyo 1971). GRH sucks sap from both the xylem and the phloem of susceptible rice varieties, leading to yield losses in northeastern Japan (Nirei and Nakazato 1975). In addition to causing direct plant damage, GRH transmits viral diseases, such as rice dwarf (Brar et al. 2009) and waika viruses, both of which are common in western Japan (Nakasuji and Nomura 1968). The level of resistance to GRH differs among rice varieties and growth stages (Kishino and Ando 1979). In resistant varieties such as IR24, Lepe-dumai, and Rantaj-emas 2, GRH resistance is strong at the young seedling and booting stages but is weaker at the tiller emergence and other growth stages. Knowledge of such differences may be useful for understanding the genetic mechanisms of GRH resistance (Fujita et al. 2010b).
Several studies to identify genes or quantitative trait loci (QTLs) that confer resistance to GLH and GRH have been reported. Genetic mapping identified a single dominant gene for resistance to GLH on chromosome 4 derived from a cross between the varieties ARC11554 and TN1 (Sebastian et al. 1996). Brar and Khush (2002) reported GLH-resistant accessions of the wild species Oryza barthii. A total of six genes and two QTLs conferring GRH resistance have been identified and mapped on seven rice chromosomes using O. sativa landraces and other Oryza species. The resistance genes include Grh1 on chromosome 5 derived from IR24, Pe-bi-hun, Singwang, and ASD7 (Kadowaki et al. 2003, Mai et al. 2015, Park et al. 2013, Tamura et al. 1999); Grh2 on chromosome 11 and Grh4 on chromosome 3 from Lepe-dumai and DV85 (Fukuta et al. 1998, Kadowaki et al. 2003); and Grh3 on chromosome 6 from Rantaj-emas 2 and Cheongnam (Hur et al. 2015, Saka et al. 2006). One major gene conferring GRH resistance from the Surinam cultivar SML17 and another one from a wild accession of Oryza nivara were mapped on the short arm of chromosome 4 and named Grh6 (Tamura et al. 2004) and Grh6-nivara (Fujita et al. 2004), respectively. A wild accession of Oryza rufipogon Griff., W1962, is highly resistant to GRH. One gene, Grh5 on chromosome 8, and a minor QTL, qGRH4, have been mapped using a backcross population between the O. sativa ssp. japonica cultivar “Taichung 65” and W1962 (Fujita et al. 2006, 2010a). Genetic resources of rice landraces were characterized (Mai et al. 2017), but the genetic analyses of GRH resistance are limited. A significant major QTL on the long arm of chromosome 9 (qGRH9) is derived from the African rice species Oryza glaberrima (Fujita et al. 2010b); however, the genetic basis of GRH resistance of O. glaberrima is not completely understood. No genes or QTLs for GRH resistance from another African rice species, Oryza longistaminata A. Chev. & Roehrich, have previously been reported.
African wild rice species are valuable sources of genetic and allelic diversity (Brar and Khush 2002, Jena 2010, Khush et al. 1990, Wambugu et al. 2013). Among the six wild species carrying the AA genome, O. longistaminata originates from tropical regions of Africa; it is a perennial species characterized by long anthers, strong rhizomes, and tolerance to blight and drought (Jena 2010, Liu et al. 2004, Neelam et al. 2018, Song et al. 1995). Mapping of QTLs for agronomically important traits such as vigorous biomass under low-input conditions and yield-enhancing traits (Gichuhi et al. 2016), QTLs for high ability to use nitrogen efficiently (Yang et al. 2010), rhizomatous expression trait (Rhz2 and Rhz3) and resistance of bacterial blight (Xa21) (He et al. 2014, Hu et al. 2003, Khush et al. 1990, Sacks et al. 2003), a novel QTL (qSPP2.2) for the number of spikelets per panicle (Kaur et al. 2018), and QTLs for seed vigor-related traits (Jin et al. 2018) has been performed using several mapping populations derived from a cross between accessions of O. longistaminata and O. sativa. Rawal et al. (2018) reported that O. longistaminata has abundant nucleotide-binding site (NBS)-encoding genes, which are important for plant defense. However, unlike with the other AA species, it is difficult to obtain F1 interspecific progeny using O. longistaminata because of embryo abortion (Causse et al. 1994, Chen et al. 2009, Sacks et al. 2003).
Development of chromosome segment substitution lines (Ramos et al. 2016) and recombinant inbred lines carrying O. longistaminata segments have been developed for QTL analysis of agronomic traits (Gichuhi et al. 2012). Introgression lines (ILs), each carrying a particular chromosome segment from the donor parent in the genetic background of the recurrent parent, have been used to identify genes for quantitative traits by fine QTL mapping (Chetelat et al. 1995, Chetelat and Meglic 2000, Eshed and Zamir 1994, 1995, 1996, Kubo et al. 2002, Yano 2001). The use of ILs for mapping of small-effect QTLs and map-based cloning has been reported (Ebitani et al. 2005, Fukuoka et al. 2010). Several sets of ILs carrying chromosome segments from wild rice in the genetic background of cultivated rice have been developed to detect QTLs for valuable traits (Ahn et al. 2002, Doi et al. 1997, Hirabayashi et al. 2010, Kurakazu et al. 2001, Sobrizal et al. 1999, Tian et al. 2006, Yoshimura et al. 2010).
To understand the genetic basis of GRH resistance derived from O. longistaminata, we developed ILs derived from O. longistaminata in the genetic background of O. sativa spp. japonica cv. ‘Nipponbare’ and detected four QTLs for resistance to GRH.
F1 plants were derived from a cross between O. sativa ssp. japonica cv. ‘Nipponbare’ as the female parent and African wild rice O. longistaminata accession W1413, originally collected from Sierra Leone, as the male parent (Fig. 1). F1 plants were successively backcrossed using Nipponbare pollen to the BC2F1 generation, and whole-genome genotyping was conducted with simple sequence repeat (SSR) markers distributed across all 12 chromosomes. BC2F1 plants were selected on the basis of their graphical genotypes; BC2F2 plants were backcrossed with Nipponbare to generate BC3F1 plants. BC3F1 plants with heterozygous for W1413 alleles in the target region of each chromosome were selected to generate progeny. A series of ILs was produced by self-pollination of BC3F2 plants to fix homozygous W1413 alleles in the target introgressed segments in the Nipponbare genetic background. These ILs were used to evaluate GRH resistance; subsequently, BC3F3 and BC3F4 populations were used in QTL analysis.
Development of ILs derived from W1413 (O. longistaminata) in the Nipponbare (O. sativa) genetic background. MAS represents marker-assisted selection. NILs mean near-isogenic lines. PYLs mean pyramided lines.
BC3F3 populations derived from individual plants of three ILs (BC3F2) and three sister lines of these ILs were used for QTL analysis. The parental BC3F2 plants carried heterozygous segments on chromosomes 1, 2, 3, 4, 5, 7, and 11 and W1413-homozygous segments on chromosomes 1, 2, 5 and 11. Seven BC3F4 populations derived from individual plants of four BC3F3 mapping populations were used for QTL validation. The parents of these four BC3F3 plants carried heterozygous segments on chromosomes 2, 4, 5, and 11 and W1413-homozygous segments on chromosomes 2, 5, and 11 in the Nipponbare genetic background.
Evaluation of GRH resistanceThe GRH strain was collected in Fukuoka Prefecture, Japan, in 1991 and was maintained by continuous rearing on seedlings of Nipponbare, which is susceptible. GRH was kept at 25 ± 1°C and 16 h light: 8 h dark. An antibiosis test for GRH resistance was conducted according to Kishino and Ando (1978) with modifications. Seedlings (10 individuals per IL, six BC3F3 populations, and seven BC3F4 populations) were grown for 7 days and infested with 10 first-instar nymphs in test tubes. Because it was difficult to obtain seeds of W1413 from self-pollination, this accession was maintained as ratoon plants, and young leaves were used for evaluation of resistance. Nymph mortality (NM, %) was calculated at 3 days after infestation (DAI), 5 DAI, and 7 DAI. Seedlings (20 per line) of four pre-NILs (near-isogenic lines), each carrying one QTL, and four pre-PYLs (pyramided lines), each carrying two or four QTLs, were also analyzed to measure the GRH resistance level. Plants were categorized as susceptible (NM <30%), moderately resistant (NM 30%–70%), or highly resistant (NM >70%).
Whole-genome genotyping and construction of a linkage mapTotal genomic DNA was extracted from freeze-dried leaves as in Dellaporta et al. (1983). The genotypes of SSR loci were determined by PCR amplification in a PCR System 9700 (PerkinElmer, Waltham, MA, USA). Primer sequences of SSR markers are listed in Supplemental Table 1. PCR mixtures (15 μl) contained 50 mM KCl, 10 mM Tris HCl (pH 9.0), 1.5 mM MgCl2, 200 μM each dNTP, 0.2 μM primer, 0.5 unit of Taq polymerase (Takara, Shiga, Japan), and 25 ng of genomic DNA. The thermal cycler was programmed as follows: 5 min at 95°C; then 40 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s. PCR products were separated by electrophoresis in 4% agarose gels at 250 V for 1 h in 0.5× TBE buffer. Gels were stained with ethidium bromide and photographed under ultraviolet light.
A genetic linkage map was constructed with 108 polymorphic SSR markers covering all chromosomes using 115 BC2F1 plants. MAPMAKER/EXP 3.0 (Lander et al. 1987) was used to determine the linkage loci and map distances.
QTL mapping and statistical analysesThe phenotype and genotype data of BC3F3 and BC3F4 plants were used for standard interval mapping with R/qtl software (Broman and Sen 2009). The critical experiment-wise threshold values of the logarithm of odds (LOD) scores for QTL analysis were calculated by conducting 1000 permutation tests at the significance level of 5%. To assess the average NM of pre-NILs and pre-PYLs, differences between means were statistically analyzed by using Scheffe’s method following analysis of variance at significance level P < 0.05.
Small leaves of ratoon plants of the W1413 accession of O. longistaminata, which were similar to those of seedlings at about 10 days after sowing (DAS), were resistant to GRH. The leaves of W1413 were also resistant at 30 DAS (pre-tillering stage) and 60 DAS (tillering stage), whereas those of Nipponbare were susceptible to GRH at all three stages. At 10 and 30 DAS, average NM of W1413 was 100% ± 0.0% (mean ± SE; N = 5) and that of Nipponbare was 0.0% ± 0.0%. At 60 DAS, it was 93.5% ± 3.0% in W1413, 1.7% ± 3.7% in Nipponbare, and 69.8% ± 4.0% in F1 plants, which thus were moderately resistant to GRH.
Construction of a series of ILs covering the whole genome of O. longistaminataF1 individuals derived from a cross between Nipponbare (female parent) and W1413 (male parent) were backcrossed with Nipponbare, and 39 BC1F1 plants were developed (Fig. 1). They were backcrossed with Nipponbare, and 115 BC2F1 plants were developed. Graphical genotypes of the BC2F1 plants showed the presence of target heterozygous introgressed segments in all 12 chromosomes. A total of 18 BC2F2 lines (36 plants per line) were grown, and 6 plants per line were backcrossed with Nipponbare. Consequently, 328 BC3F1 plants were developed and genotyped with 108 SSR markers (Supplemental Table 1); of these, 27 BC3F1 lines (36 plants per line) that carried target heterozygous segments on each chromosome were self-pollinated to produce BC3F2 populations. Candidate BC3F3 plants (26 ILs from each BC3F2 plant and 2 ILs derived from the two sister plants of BC3F2 5-13 and 5-28) were chosen; they were homozygous for W1413 alleles in the target introgressed regions, had heterozygous segments in some chromosomes (fixation of Nipponbare alleles in the next generation) and were homozygous for Nipponbare alleles in most of the non-target regions of each chromosome. Graphical genotypes of these 28 ILs were determined using the 108 SSR markers (Fig. 2). The entire genome of O. longistaminata was covered except for 11 regions located on chromosomes 1, 3, 4, 6, 7, 11, and 12 that were homozygous for Nipponbare alleles, because the corresponding chromosomal segments from W1413 were lost prior to MAS in the BC2F1 generation.
Graphical genotypes of 28 ILs derived from a cross between Nipponbare and W1413. Chr. means chromosome. Red and yellow boxes indicate homozygous for W1413 and heterozygous segments, respectively. The missing genotypes are shown in grey.
We evaluated the resistance of the 28 BC3F3 ILs (Fig. 3). Four BC3F3 ILs had moderate to high resistance at 7 DAI: BC3F3 3 (0%–50% NM), BC3F3 9 (0%–90% NM), BC3F3 21 (0%–85.7% NM), and BC3F3 26 (0%–85.7% NM). IL 3 (BC3F3 3) carried the segments of W1413 on chromosomes 1 and 5. IL 9 (BC3F3 9) was homozygous for the target W1413 segment on chromosome 2 with heterozygous segments on chromosomes 4, 5, and 11. IL 26 (BC3F3 26) carried the target segment on chromosome 11 with small heterozygous segments on chromosomes 4 and 7 (Fig. 2). The number of seeds of BC3F3 21 was limited because of high sterility. The other 24 ILs had NM from 0% to 30%.
Scatter plot of nymph mortality (%) of 28 ILs derived from O. longistaminata in the Nipponbare genetic background. S, MR and HR represent susceptible, moderately resistant, and highly resistant classes, respectively. NM indicates nymph mortality (%). ILs segregating for moderate to high resistance is encircled. Black diamonds mean average values of nymph mortality (%). Error bar shows standard error.
A total of six BC3F3 populations derived from the candidate BC3F2 plants, namely BC3F3 3 from BC3F2 3-6 (Fig. 4a), BC3F3 103 from BC3F2 3-4 (Fig. 4b), BC3F3 9 from BC3F2 7-11 (Fig. 4c), BC3F3 107 from BC3F2 7-35 (Fig. 4d), BC3F3 26 from BC3F2 25-4 (Fig. 4e), and BC3F3 123 from BC3F2 25-34 (Fig. 4f), were subjected to QTL analysis. The graphical genotypes of the six candidate BC3F2 individuals are shown in Fig. 4.
Graphical genotypes of six individual BC3F2 plants derived from a cross between Nipponbare and W1413, and frequency distributions of nymph mortality (%) at 7 DAI in the six BC3F3 populations derived from these plants. The names of the BC3F2 plant and corresponding BC3F3 population are indicated in the middle of each panel (a–f). Chromosome number is indicated above each chromosome. Black and white rectangles indicate homozygous for W1413 and Nipponbare segments, respectively. SSR markers are listed to the right of each chromosome. Black and white arrowheads indicate average nymph mortality in Nipponbare and W1413, respectively.
The BC3F2 3-6 plant carried W1413-homozygous segments on chromosomes 1 and 5 (Fig. 4a). Among a total of 90 BC3F3 3 individuals, continuous distribution from low to high NM at 7 DAI was observed (Fig. 4a). However, all plants of BC3F3 population 3 had fixed homozygous W1413 alleles on chromosomes 1 and 5; thus, the cause of segregation of NM in this population was unknown and it was excluded from QTL analysis.
BC3F2 3-4, a sister plant of BC3F2 3-6, carried heterozygous segments on chromosomes 1, 3, and 5 (Fig. 4b). Among a total of 76 BC3F3 103 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 4b). A significant QTL (qGRH5) was detected between RM509 and RM430 on chromosome 5 in the BC3F3 103 population and explained 12.8% of phenotypic variation (PV) with an additive NM value of 21.6 (Table 1). No significant QTLs on chromosomes 1 and 3 were detected at the 5% significance level.
| Population | QTL | Chromosome | Marker interval | Peak LOD | PVE (%) | Additive effecta | Dominance effect | LOD thresholdb |
|---|---|---|---|---|---|---|---|---|
| BC3F3 103 | qGRH5 | 5 | RM509–RM430 | 2.1 | 12.8 | 21.6 | 4.7 | 2.1 |
| BC3F3 9 | qGRH11 | 11 | RM5960–RM6680 | 2.0 | 11.0 | 14.6 | 15.9 | 1.8 |
| BC3F3 107 | qGRH4 | 4 | RM5414–RM8213 | 3.0 | 22.0 | 23.9 | 8.0 | 2.3 |
| BC3F3 26 | qGRH4 | 4 | RM5414–RM8213 | 1.3 | 7.1 | 23.0 | 2.1 | 1.3 |
| BC3F3 123 | qGRH11 | 11 | RM5960–RM6680 | 2.6 | 11.2 | 16.8 | 5.8 | 2.1 |
PVE represents phenotypic variation explained.
The BC3F2 7-11 plant carried heterozygous segments on chromosomes 4, 5, and 11, and a W1413-homozygous segment on chromosome 2 (Fig. 4c). Among a total of 79 BC3F3 9 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 4c). A significant QTL (qGRH11) was detected between RM5960 and RM6680 on chromosome 11 and explained 11.0% of PV with an additive NM value of 14.6 (Table 1). No significant QTLs on chromosomes 4 and 5 were detected at the 5% significance level in this population.
BC3F2 7-35, a sister plant of BC3F2 7-11, carried heterozygous segments on chromosomes 2, 4, and a W1413-homozygous segment on chromosomes 5 and 11 (Fig. 4d). Among a total of 55 BC3F3 107 individuals, a pseudo-normal distribution from low to high NM at 7 DAI was observed (Fig. 4d). A significant QTL (qGRH4) was detected between RM5414 and RM8213 and explained 22.0% of PV with an additive NM value of 23.9 (Table 1). No significant QTLs on chromosomes 2 and 5 were detected at the 5% significance level.
The BC3F2 25-4 plant carried heterozygous segments on chromosomes 4 and 7, and a W1413-homozygous segment on chromosome 11 (Fig. 4e). Among a total of 84 BC3F3 26 individuals, a continuous distribution from low to high NM was observed (Fig. 4e). One significant QTL (qGRH4 described above) was detected and explained 7.1% of PV with an additive NM value of 23.0 (Table 1). No significant QTLs on chromosome 7 were detected at the 5% significance level.
BC3F2 25-34, a sister plant of BC3F2 25-4, carried heterozygous segments on chromosomes 4, 7, and 11 (Fig. 4f). Among a total of 99 BC3F3 123 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 4f). A significant QTL (qGRH11 described above) was detected in the BC3F3 123 population and explained 11.2% of PV with an additive NM value of 16.8 (Table 1). No significant QTLs on chromosomes 4 and 7 were detected at the 5% significance level. Thus, three significant QTLs, qGRH4, qGRH5, and qGRH11, were identified in the five BC3F3 populations (Table 1).
Validation of QTLs for GRH resistanceSeven BC3F3 individuals were selected from four BC3F3 mapping populations on the basis of the genotypes of candidate QTL regions. The graphical genotypes of the selected individuals are shown in Fig. 5.
Graphical genotypes of individual BC3F3 plants derived from a cross between Nipponbare and W1413, and frequency distributions of nymph mortality (%) at 7 DAI in the seven BC3F4 populations derived from these plants. The names of the BC3F3 plant and corresponding BC3F4 population are indicated in the middle of each panel (a–f). Chromosome number is indicated above each chromosome. Black and white rectangles indicate homozygous for W1413 and Nipponbare segments, respectively. Detected QTL regions are shown in ovals. SSR markers are listed to the right of each chromosome. Markers linked to QTLs are underlined. Black and white arrowheads indicate average nymph mortality in Nipponbare and W1413, respectively. In frequency distribution, white, black and gray colors show the Nipponbare homozygous, W1413 homozygous and heterozygous at RM5414 (5a), RM509 (5b, 5c), RM5960 (5d), RM3688 (5e, 5f), respectively.
The BC3F3 123-47 plant was selected to validate qGRH4 (Fig. 5a). Among a total of 93 BC3F4 4 individuals, bimodal segregation with low and high NM at 7 DAI was observed (Fig. 5a). Only one QTL (qGRH4) was detected in the BC3F4 4 population and explained 14.8% of PV with an additive NM value of 13.7 (Table 2).
| QTLs | Chromosome | Marker interval | Peak LOD | PVE (%) | Additive effecta | Dominance effect | Population | LOD thresholdb |
|---|---|---|---|---|---|---|---|---|
| qGRH4 | 4 | RM5414–RM8213 | 3.2 | 14.8 | 13.7 | 10.8 | BC3F4 4 | 1.5 |
| qGRH5 | 5 | RM509–RM430 | 2.7 | 12.3 | 12.6 | 8.5 | BC3F4 9 | 1.6 |
| qGRH5 | 5 | RM3381–RM509 | 4.2 | 19.2 | 13.9 | 10.7 | BC3F4 15 | 1.8 |
| qGRH11 | 11 | RM5960–RM6680 | 5.2 | 21.5 | 15.2 | 10.8 | BC3F4 1 | 1.5 |
| qGRH2 | 2 | RM3688–RM3515 | 2.7 | 12.6 | 17.7 | 1.9 | BC3F4 7 | 1.8 |
| qGRH4 | 4 | RM5414–RM8213 | 3.7 | 16.7 | 13.4 | 6.7 | BC3F4 12 | 1.8 |
| qGRH2 | 2 | RM3688–RM3515 | 3.3 | 15.0 | 13.2 | 0.2 | BC3F4 12 | 1.8 |
PVE represents phenotypic variation explained.
The BC3F3 103-41 plant was selected to validate qGRH5 (Fig. 5b). Among a total of 95 BC3F4 9 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 5b). Only one QTL (qGRH5) was detected in the BC3F4 9 population and explained 12.3% of PV with an additive NM value of 12.6 (Table 2).
The BC3F3 9-76 plant was also selected to validate qGRH5; this plant had a W1413-homozygous segment on chromosome 2 (Fig. 5c). Among a total of 90 BC3F4 15 individuals, a continuous distribution from low to high NM was observed (Fig. 5c). Only one QTL (qGRH5) was detected in the BC3F4 15 population and explained 19.2% of PV with an additive NM value of 13.9 (Table 2).
The BC3F3 123-48 plant was selected to validate qGRH11 (Fig. 5d). Among a total of 93 BC3F4 1 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 5d). Only one major QTL (qGRH11) was detected in the BC3F4 1 population and explained 21.5% of PV with an additive NM value of 15.2 (Table 2).
In addition, the BC3F3 9-60 plant, which carried heterozygous segments at qGRH4, qGRH5, and qGRH11, and a W1413-homozygous segment on chromosome 2 (graphical genotype identical to that of BC3F2 7-11; Fig. 4c), was selected to validate candidate QTLs. Among a total of 67 BC3F4 2 individuals, a continuous distribution from low to high NM at 7 DAI was observed (data not shown). Only one major QTL (qGRH11) was detected in the BC3F4 2 population and explained 21.1% of PV with an additive NM value of 18.1 (data not shown). No significant QTLs on chromosomes 4 and 5 were detected at the 5% significance level in this population. Thus, only qGRH11 was detected both in the BC3F4 2 and BC3F3 9 populations (Table 1).
The BC3F3 107-8 plant, which carried a heterozygous segment on chromosome 2 and W1413-homozygous segments at qGRH5 and qGRH11 (Fig. 5e), was selected to assess whether a QTL is present on chromosome 2. Among a total of 93 BC3F4 7 individuals, a continuous distribution from low to high NM at 7 DAI was observed (Fig. 5e). A novel QTL (qGRH2) was detected between RM3688 and RM3515 on the long arm of chromosome 2 in the BC3F4 7 population. qGRH2 explained 12.6% of PV with an additive NM value of 17.7 (Table 2).
The BC3F3 107-24 plant, which carried heterozygous segments on chromosome 2 and at qGRH4, and W1413-homozygous segments at qGRH5 and qGRH11 (Fig. 5f), was selected for validating qGRH2 and qGRH4. Among a total of 92 BC3F4 12 individuals, a continuous distribution from low to high NM was observed (Fig. 5f). Two QTLs (qGRH2 and qGRH4) were detected in the BC3F4 12 population. The location of qGRH2 between RM3688 and RM3515 was confirmed; qGRH2 explained 15.0% of PV with an additive NM value of 13.2. The QTL qGRH4 explained 16.7% of PV with an additive NM value of 13.4 (Table 2).
Nymph mortality of pre-NILs and pre-PYLsFour pre-NILs with W1413-homozygous alleles on the long arm of chromosome 2, the short arm of chromosome 4, the short arm of chromosome 5, and the long arm of chromosome 11 and four pre-PYLs (qGRH2/qGRH11-PYL, qGRH4/qGRH11-PYL, qGRH5/qGRH11-PYL, and qGRH2/qGRH4/qGRH5/qGRH11-PYL) were selected from BC3F3 mapping populations and evaluated (N = 20) for GRH resistance (Fig. 6). The young leaves of W1413 ratoon plants were used as the resistant control. At 3 DAI, NM did not differ among the pre-NILs and pre-PYLs, but was significantly lower than that in W1413 (Fig. 6a). At 5 DAI, the NM increased but did not differ significantly among the pre-NILs and pre-PYLs except for qGRH5/qGRH11-PYL, which had the lowest NM, and the pre-PYL carrying four QTLs, which had the highest. The NM of all pre-NILs and pre-PYLs was significantly lower than that of W1413 (Fig. 6b). At 7 DAI, the NM levels of pre-NILs carrying qGRH2, qGRH4, and qGRH5 did not differ significantly from each other, but were significantly lower than those of qGRH11-NIL, the pre-PYLs, and W1413. The NM of the pre-NIL carrying qGRH11 did not differ significantly from those of the pre-PYLs but was lower than those of the four-QTL pre-PYL and W1413. Among all the lines, the pre-PYL carrying four QTLs showed the highest resistance level (97.3% ± 1.3), which was equivalent to that of W1413 (100.0% ± 0.0) (Fig. 6c).
Nymph mortality (NM) on pre-NILs, pre-PYLs, W1413, and Nipponbare at different days after infestation (DAI). Means denoted with different letters are significantly different at the 5% level according to Scheffe’s method. Error bars indicate standard error. W1413 leaves were from ratoon plants.
After fixing W1413 introgression segments on target chromosomes in the Nipponbare genetic background, a series of 28 ILs was evaluated for GRH resistance. Among them, four candidates had moderate to high NM. Candidate IL populations and populations derived from sister plants carrying heterozygous segments were used in QTL analysis. Segregating populations carrying chromosomal segments from W1413 in the Nipponbare genetic background allowed us to map QTLs for GRH resistance from the wild rice species O. longistaminata (Tables 1, 2).
Resistance levels of O. longistaminata and its derivativesThe donor parent, accession W1413, was highly resistant to GRH (up to NM 100.0% ± 0.0% at some stages) when tested by using leaves from ratoon plants. None of the segregating populations contained plants resistant at 3 DAI (Supplemental Figs. 1, 3); however, the NM at 5 and 7 DAI were distributed continuously from susceptibility to resistance (Figs. 4, 5, Supplemental Figs. 2, 4), indicating that multiple additive QTLs control GRH resistance derived from W1413.
Four significant QTLs for GRH resistanceFour significant QTLs (qGRH2, qGRH4, qGRH5, and qGRH11) were validated by using 12 advanced backcross populations (Tables 1, 2). The QTLs qGRH4, qGRH5, and qGRH11 were located on chromosomes 4, 5, and 11, respectively, in similar locations to the previously reported genes Grh6, Grh1, and Grh2, respectively. qGRH4, in the distal region of chromosome 4, was detected between the RM5414 and RM8213 markers (Fig. 5a, 5f); the Grh6-nivara gene, on the short arm of chromosome 4, is tightly linked with RM8213 and is located between the RM5414 and C60248 SSR markers (Fujita et al. 2004). qGRH5 had an LOD peak between the RM3381 and RM509 markers (Table 2, Fig. 5b, 5c), and the Grh1 gene from Pe-bi-hun and IR24 rice landraces is tightly linked with RM3381 (Kadowaki et al. 2003, Tamura et al. 1999, Yasui and Yoshimura 1999). qGRH11, on the long arm of chromosome 11, was similar to a candidate region of the previously mapped Grh2 gene from Lepe-dumai and DV85 rice landraces (Fukuta et al. 1998, Kadowaki et al. 2003, Yazawa et al. 1998). The LOD peak of qGRH11 was detected between the RM5960 and RM6680 markers (Fig. 5d), and qGRH11 is tightly linked with RM5960; the Grh2 gene is tightly linked with the RM5961 marker (Fujita et al. 2010a). A QTL on chromosome 2, qGRH2, was detected between RM3688 and RM3515 (Table 2) in the BC3F4 7 and BC3F4 12 populations, which were W1413-homozygous at qGRH5 and qGRH11 in the Nipponbare genetic background. qGRH2 has not been previously reported.
Genetic effects of the detected QTLs (qGRH4, qGRH5, and qGRH11) from W1413 were identified in BC3F4 populations 4, 9, and 1, respectively, all with the Nipponbare genetic background (Table 2). The NM distribution of BC3F4 4 showed about 50% of the plants was susceptible to GRH (≤30% NM) (Fig. 5a). One QTL, qGRH4, was detected in this population (Table 2). In BC3F4 9, about 50% of the plants were also susceptible to GRH (≤30% NM) (Fig. 5b). qGRH5 was identified in this population (Table 2). In BC3F4 1, about 35% of the plants had susceptible to GRH (≤30% NM) (Fig. 5d). A major QTL, qGRH11, was detected in this population (Fig. 5d, Table 2). No plant materials for detection of qGRH2 (heterozygous only on chromosome 2) in the Nipponbare genetic background were available in BC3F3 mapping populations, but it was detected as a significant QTL in the background of qGRH5 and qGRH11 in BC3F4 7 (Fig. 5e, Table 2).
QTL analysis of BC3F3 populations detected qGRH4 in BC3F3 107 and 26, which were W1413-homozygous at qGRH11 (Fig. 4d, 4e, Table 1). However, qGRH4 was never detected as a significant QTL when qGRH11 was heterozygous (in BC3F3 9 and 123; Fig. 4c, 4f, Table 1). These findings suggest that the major effect of qGRH11 masked those of the minor QTLs.
In BC3F4 2, qGRH11 was identified as a major QTL (high PV explained [PVE] and high additive value) in the presence of heterozygous qGRH4 and qGRH5 regions and W1413-homozygous qGRH2 (data not shown). About 40% of BC3F4 2 plants and 25% of BC3F4 1 plants (containing only qGRH11) were highly resistant (≥60% NM) (Fig. 5d). These data suggest that a combination of qGRH11 with other minor QTLs increases resistance; the number of individual plants (N = 67) in BC3F4 2 was insufficient to detect the individual effect of the minor QTLs (data not shown). The BC3F4 15 population, in which we detected qGRH5 and which was W1413-homozygous at qGRH2, had an increased proportion of plants with high resistance (≥60% NM) (Fig. 5c); in this population, qGRH5 had a high PVE and increased additive value (Table 2). BC3F4 9, in which we detected qGRH5 in the absence of other significant QTLs, had an increased proportion of susceptible plants (≤30% NM) (Fig. 5b). These results indicate a combined effect of two additive QTLs, qGRH5 and qGRH2, in BC3F4 15.
Among the detected QTLs, qGRH11 from W1413 is a major QTL with the highest PVE (21.5%) in the BC3F4 1 population (Table 2). The other minor QTLs are qGRH4 (PVE = 14.8%–16.7%) in BC3F4 4 and 12, qGRH5 (PVE = 12.3%–19.2%) in BC3F4 9 and 15, and qGRH2 (PVE = 12.6%−15.0%) in BC3F4 7 and 12 (Table 2).
Pyramiding of four resistance allelesBC3F4 12 had an increased proportion of plants with moderate to high resistance (≥50% NM), and 25% of plants were susceptible to GRH (Fig. 5f). Two QTLs, qGRH2 and qGRH4, were detected in the presence of W1413- homozygous qGRH5 and qGRH11 in BC3F4 12 (Table 2). Therefore, the combined effect of the four QTLs from W1413 conferred high resistance to GRH.
Pre-NILs and pre-PYLs were evaluated to ascertain the effect of each QTL and of various combinations (Fig. 6). At 3 DAI, the NM of all lines was significantly lower than that of W1413, with no differences among the lines. At 5 DAI, the NM of all pre-NILs and pre-PYLs gradually increased, and the resistance of the pre-PYL carrying four QTLs was significantly greater than that of other lines. At 7 DAI, the NM of all lines increased further, and that of the pre-PYL carrying four QTLs did not differ significantly from that of W1413. Thus, the resistance level of W1413 at 7 DAI can be largely explained by the combined effects of the four resistance alleles (qGRH2, qGRH4, qGRH5, and qGRH11). Otherwise, evaluation of W1413 young ratoon leaves expressed the resistance of additional QTLs besides from these four alleles and additional QTLs might not be detected in this study.
In conclusion, we developed a series of ILs from a cross between Nipponbare and W1413, and mapped four GRH resistance QTLs using advanced backcross populations. The resistance of the pre-NILs, and pre-PYLs increased gradually with time: all lines were susceptible at 3 DAI, but became moderately resistant at 5 DAI and highly resistant at 7 DAI. Our results suggest that multiple QTLs control GRH resistance in W1413. The chromosomal locations of qGRH4, qGRH5, and qGRH11 were identical to those of previously reported GRH resistance genes Grh6, Grh1, and Grh2, respectively. In addition, we identified qGRH2, a new QTL for GRH resistance. The genetic effect of qGRH11 was the largest and the other QTLs had minor effects. We conclude that at least these four QTLs contribute to the resistance to GRH derived from O. longistaminata.
HWT and HY designed the experiments. HWT, YY, MVT, and HY developed plant materials. HWT conducted the genetic analysis. HWT and HY wrote the paper.
We are grateful to the Biotron Application Center, Kyushu University, for the use of the Entomotron in our antibiosis tests. This work was supported by a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan [Genomics-based Technology for Agricultural Improvement (LTC-0010)]. This work was partly supported by grants from the Science and Technology Research Partnership for Sustainable Development (SATREPS)–Myanmar established by Japan Science and Technology Agency/Japan International Cooperation Agency (JST/JICA). We thank the Ministry of Education, Culture, Sports, Science and Technology (MEXT) for a scholarship granted to HWT and SATREPS–Vietnam by JST/JICA for a doctoral fellowship granted to TVM. Accessions used in this study were obtained from the National Institute of Genetics supported by the National Bioresource Project, MEXT, Japan.
